Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Use common track-TP association in track and vertex validation #8888

Conversation

makortel
Copy link
Contributor

Previously tracks and TrackingParticles were associated to each other separately in track (all tracks + once for each iteration, and x2 for high purity tracks) and in vertex validation. In this PR the production of these track-TP associations is consolidated (the gain in time is, however, small as making the track-TP association in the presence of ClusterTPAssociationProducer is relatively fast).

In order to continue to support the per-iteration efficiency plots, a generic helper function (associationMapFilterValues) is implemented to enable filtering key-value pairs from AssociationMap based on the value. This allows to do the track-TP associations once with full collections, and then for iteration-specific easily restrict the TP-track association to the tracks of that iteration.

The configuration interface of MultiTrackValidator is slightly modified to use the same parameter for delivering the InputTags for associations/associators. The flag UseAssociators dictates whether the associators (True) are used or the associations (False). All other uses than validation for pp (as found with git grep) are migrated.

In addition, I propagated the change from iteration numbers to names (#5807) to the labels of the iteration-specific track selectors. Since the names of the MTV histogram directories change anyway in this PR, this way we get these two renames of the directories into a single validation cycle (for sure).

Tested in CMSSW_7_5_X_2015-04-12-2300, but rebased on top of CMSSW_7_5_X_2015-04-21-2300 due to new configuration files in FastSim/Validation that had to be migrated. No changes expected in histogram contents.

@rovere @VinInn

Now both associator and association map are given as InputTags via
VInputTag. This is a preparatory step for unifying the associations
between MTV and vertex validation.
The FastSim way of doing the track-TP association is without the
cluster-TP association. If the cluster-TP associator is accidentally
introduced to the FastSim validation sequence,
QuickTrackAssociatorByHits uses that and gives 0 associations. It is
better to explicitly set QuickTrackAssociatorByHits to not to use
cluster-TP association.
@makortel
Copy link
Contributor Author

I have here https://mkortela.web.cern.ch/mkortela/tracking/pr8888/compareValHists.C a version of compareValHists.C used in alternative-comparisons properly mapping the old and new directory names so that the histograms can be compared (although manually).

@cmsbuild
Copy link
Contributor

A new Pull Request was created by @makortel (Matti Kortelainen) for CMSSW_7_5_X.

Use common track-TP association in track and vertex validation

It involves the following packages:

CommonTools/Utils
FastSimulation/Validation
SLHCUpgradeSimulations/Geometry
Validation/Configuration
Validation/RecoHI
Validation/RecoTrack
Validation/RecoVertex

@civanch, @Dr15Jones, @lveldere, @cvuosalo, @ianna, @mdhildreth, @monttj, @cmsbuild, @ssekmen, @nclopezo, @deguio, @slava77, @vadler, @ktf, @danduggan can you please review it and eventually sign? Thanks.
@Martin-Grunewald, @kkrajczar, @GiacomoSguazzoni, @rovere, @VinInn, @appeltel, @richard-cms, @matt-komm, @RylanC24, @wmtford, @cerati, @dgulhan, @venturia this is something you requested to watch as well.
You can sign-off by replying to this message having '+1' in the first line of your reply.
You can reject by replying to this message having '-1' in the first line of your reply.
If you are a L2 or a release manager you can ask for tests by saying 'please test' in the first line of a comment.
@nclopezo you are the release manager for this.
You can merge this pull request by typing 'merge' in the first line of your comment.

@slava77
Copy link
Contributor

slava77 commented Apr 27, 2015

@cmsbuild please test

@cmsbuild
Copy link
Contributor

The tests are being triggered in jenkins.

@davidlange6
Copy link
Contributor

[which seems to be for events with nothing in the hard scatter event]

@makortel
Copy link
Contributor Author

makortel commented May 8, 2015

Likely yes, I'll take a look.

@makortel
Copy link
Contributor Author

makortel commented May 8, 2015

@davidlange6 Fix is in #9000.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

Successfully merging this pull request may close these issues.

None yet

10 participants